Cultural and Natural Resources in Tourism Island: Bibliometric Mapping

: Island ecosystems have very speciﬁc physical, economic and socio-cultural characteristics, which are shared by most of these ecosystems regardless of their geographical area. These characteristics include well deﬁned geographical boundaries that lead to a greater degree of isolation, lack of economically exploitable resources, great external dependence for consumption, cultural and natural heritage with a high uniqueness level and a high concentration of endemic plant and animal species. All of them are responsible for the high development dependence level linked to the tourism activity of these ecosystems. Thus, island ecosystems are currently an important international tourism destination, where a great diversity of very attractive natural and cultural resources and of great heritage value are concentrated. This fact allows for the development of tourism activities of great heterogeneity among countries or regions of the world that host these island ecosystems. The aim of this research was to identify and analyze, through a bibliometric and bibliographic analysis, the scientiﬁc production indexed in the international Scopus database, which addresses the subject of use of cultural and natural resources by tourism in island ecosystems. This scientiﬁc mapping allows us to observe the evolution of scientiﬁc production in this ﬁeld of study. The results show that this is a new subject (a large number of transient authors), predominantly a ﬃ liated with the United States and Spain. The most followed research lines to date are destination management followed very far by responsible environmental behavior and the impacts of sustainable development. However, the keywords with the highest co-occurrence show that the hot topics are tourism exploitation in the ecotourism ﬁeld and sustainable tourism development. This research is considered the ﬁrst bibliometric study carried out which is related to this thematic approach, providing a clear in-depth analysis for researchers and thereby facilitating the approach of future research work.


Cultural and Natural Heritage
Constant debates on the relevance of the cultural and natural heritage of territories established that such assets are an irreplaceable source of identity and inspiration, which can be used as key implementing spaces that can generate other capitalization methods has also become a constant choice alternative [39,40].
In this context, this research aims to analyze and examine the scientific production developed in relation to use of cultural and natural resources by tourism in a geographical environment such as island ecosystems, in order to observe its evolution. The methodology used is a bibliometric analysis that allows for scientific mapping based on the scientific documents identified in the international Scopus database (185 documents). The novelty of this research is that there is no compilation work of the scientific production of these characteristics on the subject under study. The results and information obtained are of great value for researchers who work on the subject or begin their first steps, as well as contributing to the literature by performing a scientific mapping on the scientific production published until 2019.
This article is structured into four sections. A brief introduction in which the subject is contextualized, the novelty of the work is presented and the objective is raised. Then, the work methodology (materials and methods) is presented and the results obtained are shown in the third section. In the last section, the conclusions and limitations of the investigation are discussed.

Materials and Methods
The methodology focuses on a bibliometric analysis. This analysis applies techniques and tools that enable to identify, document and synthesize different characteristics of the field of knowledge [41,42]. Based on the data obtained, scientific mapping is carried out, outlining the networks based on their structure and definition of their relationships between the elements of the structure [43]. An appropriate articulation of these networks will identify future lines of strengthening the scientific production of this field [42].

Search Criteria and Source Identification
Firstly, the scope of the review was defined. The Scopus database was selected with the purpose of identifying high-quality scientific material based on three characteristics: (a) application of rigorous quality standards through the Relative Quality Index, Scimago Journal Rank (SJR) [44]; (b) wide coverage in time compared to other databases, being more suitable for a citation analysis because Scopus provides about 20% more coverage than its main competitor, WoS [45]; and (c) access to simultaneous downloads of a considerable number of references stored in detail as metadata [46]. The requirement and fulfilment of these characteristics allows for the validation of indexed information and therefore, its use is justified [47].
Several criteria are established for searching and monitoring information were established for developing the review. First, coverage time is determined to be equal to or less than 2019, with the purpose of identifying information published during full years. Then, the documentary unit of analysis is established, which, in this case, is the article; the choice is made based on joining the six criteria: (1) speed within publication processes, (2) rigorous arbitration process, (3) visibility and impact at different levels (local, national and international), (4) specific outline of presentation of information, (5) acceleration in positioning of authors within the scientific community, and (6) increase in feedback [48][49][50]. Conference articles, books and chapters, editorials, notes, letters or errors contained in the selected base are excluded from the process.
Tracking the selected documentary units was based on the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyzes (PRISMA) method, detailed in Figure 1 [51,52]. This methodology is composed of an evidence-based minimum set of items for reporting in systematic reviews and meta-analyses. Specifically, it consists of a 27-item checklist and a 4-phase flow diagram. The first step is to identify scientific production records through a search equation. These records contain a set of metadata that describes their attributes [53] and are the main source of information for scientometric and bibliometric studies [54]. The process begins by structuring an advanced search equation, because within Scopus, this option manages to combine the basic search options without the author limits and search, and more operators and codes are enabled [45]. The equation is constructed as follows: (TITLE-ABS-KEY (island AND touris *) AND TITLE-ABS-KEY (cultural AND resources) OR TITLE-ABS-KEY (natural AND resources)) AND DOCTYPE (ar OR re) AND PUBYEAR < 2019. This equation resulted in 482 documents for analysis.

Data Extraction
Once the documents were identified, the exclusion criteria were applied to refine and standardize the metadata by eliminating duplicate documents, those with unidentified bibliometrics and those unrelated to the subject of study. After the final selection and verification of eligibility, 185 documents were consolidated within Scopus for use in the review.
To prepare the Scopus sources for bibliometric analysis, the records of the 185 documents included in csv format (comma separated values) were downloaded. Metadata on citation information, bibliographic information, summary, keywords, among other information were included in the download. The downloaded information was worked with in the Microsoft-Office Excel software based on the analysis variables required for each selected bibliometric indicator. In addition, VOSviewer bibliometric analysis software was used for developing scientific mapping (build and visualize bibliometric networks) [55][56][57][58][59][60][61][62]. Finally, EndNote was used to manage bibliographic references.

Data Analysis
Data analysis is structured in two parts. In the first one, a descriptive analysis is developed, which details the basic characteristics of the scientific production of the field of knowledge. For this, a series of indicators that arise from different mathematical models, which are based on the relationship of two or more variables were applied to the total set of identified metadata. These are classified into activity indicators that provide information on quantity, productivity, dispersion, collaboration and networks, among others; and into impact indicators, which provide information on the citation level of the documents, impact factor or immediacy index, H-index (measure both the productivity and citation impact of the scientific publications of a researcher; takes into account the most cited articles of the researcher and the number of citations received in other publications), among others [63].
In the second part of the analysis, VOSviewer software was used, which enabled to build different representations of the scientific mapping, thus identifying the different relationships that are built in the field of study knowledge. In this section, the co-citation or citation analysis was also included, which is established according to Klavans and Boyack [64] as "the most effective way to index the literature" (p. 2). This analysis enables to determine the frequency with which other authors

Data Extraction
Once the documents were identified, the exclusion criteria were applied to refine and standardize the metadata by eliminating duplicate documents, those with unidentified bibliometrics and those unrelated to the subject of study. After the final selection and verification of eligibility, 185 documents were consolidated within Scopus for use in the review.
To prepare the Scopus sources for bibliometric analysis, the records of the 185 documents included in csv format (comma separated values) were downloaded. Metadata on citation information, bibliographic information, summary, keywords, among other information were included in the download. The downloaded information was worked with in the Microsoft-Office Excel software based on the analysis variables required for each selected bibliometric indicator. In addition, VOSviewer bibliometric analysis software was used for developing scientific mapping (build and visualize bibliometric networks) [55][56][57][58][59][60][61][62]. Finally, EndNote was used to manage bibliographic references.

Data Analysis
Data analysis is structured in two parts. In the first one, a descriptive analysis is developed, which details the basic characteristics of the scientific production of the field of knowledge. For this, a series of indicators that arise from different mathematical models, which are based on the relationship of two or more variables were applied to the total set of identified metadata. These are classified into activity indicators that provide information on quantity, productivity, dispersion, collaboration and networks, among others; and into impact indicators, which provide information on the citation level of the documents, impact factor or immediacy index, H-index (measure both the productivity and citation impact of the scientific publications of a researcher; takes into account the most cited articles of the researcher and the number of citations received in other publications), among others [63].
In the second part of the analysis, VOSviewer software was used, which enabled to build different representations of the scientific mapping, thus identifying the different relationships that are built in the field of study knowledge. In this section, the co-citation or citation analysis was also included, which is established according to Klavans and Boyack [64] as "the most effective way to index the literature" (p. 2). This analysis enables to determine the frequency with which other authors jointly cite two authors or documents in the reference list of the database under review [42,43,65,66]. It also counts the aforementioned citations, a value that enables to build similarity measures through bibliometric mapping [64,67].
The analysis of co-occurrence of keywords or co-keywords was also included. The aim of this analysis is to determine the frequency with which certain keywords "occur together" in three possible sections of the reviewed documents; title, abstract or keywords. Chen et al. [68] propose to observe the relationship generated from the presence of two or more keywords within the same research topic, so the more frequent the coincidence of these combinations is, the closer the relationship is. This analysis is carried out based on the study of the metadata of the documents registered and is classified as a method of great value for researchers who begin in new research fields, since an overview of the analyzed knowledge field can be available from a relational structure [69,70], in addition to allowing to track emerging concepts as they occur within the analyzed literature [64].

Productivity by Years
185 documents have been identified since 1985. The first work is Tourism resources and their development in Maldive Islands [71]. The production is concentrated in 92% within the last 19 years, supporting the thesis of Price's Law, the duplication of production in the course of 10 to 15 years after having begun the study of the thematic approach [72]. 2013 and 2017 are the years with the greatest production, with a concentration of 18 items each.
The evolution of scientific production is observed in Figure 2. At first, there is a slow growth (precursor stage) that extends until the year 2000; 80% of documents have a single authorship, which implies a ratio of 0.94 articles/year. As of 2001, there is a substantial increase in scientific production (accelerated growth), 73% of the articles show a participation of two or more authors per publication with a ratio of 8.95 articles/years in this period.
Sustainability 2020, 12, x FOR PEER REVIEW 6 of 27 jointly cite two authors or documents in the reference list of the database under review [42,43,65,66]. It also counts the aforementioned citations, a value that enables to build similarity measures through bibliometric mapping [64,67]. The analysis of co-occurrence of keywords or co-keywords was also included. The aim of this analysis is to determine the frequency with which certain keywords "occur together" in three possible sections of the reviewed documents; title, abstract or keywords. Chen et al. [68] propose to observe the relationship generated from the presence of two or more keywords within the same research topic, so the more frequent the coincidence of these combinations is, the closer the relationship is. This analysis is carried out based on the study of the metadata of the documents registered and is classified as a method of great value for researchers who begin in new research fields, since an overview of the analyzed knowledge field can be available from a relational structure [69,70], in addition to allowing to track emerging concepts as they occur within the analyzed literature [64].

Productivity by Years
185 documents have been identified since 1985. The first work is Tourism resources and their development in Maldive Islands [71]. The production is concentrated in 92% within the last 19 years, supporting the thesis of Price's Law, the duplication of production in the course of 10 to 15 years after having begun the study of the thematic approach [72]. 2013 and 2017 are the years with the greatest production, with a concentration of 18 items each.
The evolution of scientific production is observed in Figure 2. At first, there is a slow growth (precursor stage) that extends until the year 2000; 80% of documents have a single authorship, which implies a ratio of 0.94 articles/year. As of 2001, there is a substantial increase in scientific production (accelerated growth), 73% of the articles show a participation of two or more authors per publication with a ratio of 8.95 articles/years in this period.

Authors
491 authors are identified in the 185 documents, which implies a productivity index of 0.97 articles per author. C. León is the most productive author, with 3 articles, followed by a group of 15 authors, with 2 articles each. On the other hand, the author that accumulates the most citations is M.C. Uyarra, with 86 citations in two documents.

Authors
491 authors are identified in the 185 documents, which implies a productivity index of 0.97 articles per author. C. León is the most productive author, with 3 articles, followed by a group of 15 authors, Sustainability 2020, 12, 724 7 of 26 with 2 articles each. On the other hand, the author that accumulates the most citations is M.C. Uyarra, with 86 citations in two documents. Table 1 identifies the key and most productive authors within the thematic approach, together with their citations. Note: R = ranking; f = frequency; TC = total number of citations received for published articles; C/f = average of citations received for published articles; h-index = Hirsch's index (the maximum value of h such that the given author has published h papers that have each been cited at least h times [73]). Source: Own elaboration.
The productivity of the total number of authors can be analyzed by means of different types of processes, which enable to classify the authors based on the quantity of documents that are contributed to the thematic approach. Crane [74] proposes the existence of four groups of authors; large producers (more than 10 documents), moderate producers (between 5 and 9 documents), aspiring and transient authors (Table 2). The existence of two groups is identified. The first group "aspiring", with 2-3 documents per author, represents 3.25% of the total authors. The second largest group of "transients", with a single document, represents 96.75% of the authors, with a Transience Index (TI = [PI = 0]). It is equal to or represents the total of occasional authors that only arise once within the review and do not make any more contributions to the rest of the line of evolution. This shows that there is a high circulation of authors addressing the thematic approach.
With regard to the collaboration trend in the production of documents, 31.4% are single authorship documents and the remaining percentage, 68.6% are multiple collaboration documents. According to Berelson [75] and López López [76], the presence of several authors with different affiliations in a document shows the maturity of the thematic approach. This thematic approach is beginning its professionalization process by showing an authorship index of 2.65 authors/article, that is, the production of multiple collaboration is mainly developed from papers written by academic peers. This collaborative trend accumulates 767 of the total number of citations.

Productivity by Type of Institution and Country
This analysis enables to determine the geographical nodes of information concentration, as well as the institutions in which the production is registered. In relation to geographical production by continent (Figure 3), it is observed that the leader is Europe (72 documents), followed by Asia (46) and America (44), although the most productive country is the United States. The distribution identified on this subject, as Yoopetch and Nimsai [42] state, is not surprising, due to the persistent Anglo-American European dominance both in the publication of articles as in the publication of journals. At this point, it is necessary to indicate that production started to intensify in the last years within Oceania and Africa, due to the emergence of new tourist destinations, as well as due to increasing levels of concern for sustainable exploitation of destinations already worked within these continents. affiliations in a document shows the maturity of the thematic approach. This thematic approach is beginning its professionalization process by showing an authorship index of 2.65 authors/article, that is, the production of multiple collaboration is mainly developed from papers written by academic peers. This collaborative trend accumulates 767 of the total number of citations.

Productivity by Type of Institution and Country
This analysis enables to determine the geographical nodes of information concentration, as well as the institutions in which the production is registered. In relation to geographical production by continent (Figure 3), it is observed that the leader is Europe (72 documents), followed by Asia (46) and America (44), although the most productive country is the United States. The distribution identified on this subject, as Yoopetch and Nimsai [42] state, is not surprising, due to the persistent Anglo-American European dominance both in the publication of articles as in the publication of journals. At this point, it is necessary to indicate that production started to intensify in the last years within Oceania and Africa, due to the emergence of new tourist destinations, as well as due to increasing levels of concern for sustainable exploitation of destinations already worked within these continents. With regard to productivity by country of affiliation, the United States is the largest producer, with 76 authors, 76 authorships and 46 centers, followed by Spain with 42 authors, 45 authorships and 18 centers (Table 3). In relation to citation accumulation by country, the United States is the leader with 985 citations, followed by Canada (748) and Spain (696).  With regard to productivity by country of affiliation, the United States is the largest producer, with 76 authors, 76 authorships and 46 centers, followed by Spain with 42 authors, 45 authorships and 18 centers (Table 3). In relation to citation accumulation by country, the United States is the leader with 985 citations, followed by Canada (748) and Spain (696).  Note: R = ranking; C = centers; A = authors; As = authorships; hi% = relative frequency; TC = total number of citations received for published articles; h-index = Hirsch's index. Source: Own elaboration.
The productivity by affiliation institution shows the presence of 286 different types of affiliation centres, among which universities concentrate 71.6% (212) of affiliations and research institutes 10.8%. The presence of affiliations associated with public sector units in the different countries is highlighted, as well as several museums of great international renown. Table 4 shows the ranking of institutions that concentrate a greater number of affiliations. James Cook University (Australia) leads the ranking with 12 affiliations, followed by the University of Las Palmas de Gran Canaria, with 10 affiliations. Wadden Academy-KNAW Netherlands 6 6 Note: R = ranking; A = authors; As = authorships. Source: Own elaboration.
The collaboration networks that are generated within this thematic approach are built under two criteria: by geographical origin, where it is identified that 70.9% of the documents are carried out by academics residing within the same country and only 29.1% in collaboration with academics from different countries. The second criterion is by institutional affiliation, 100% of the documents produced in multiple collaboration at international level are by academics who are affiliated in centers located in different countries, whereas 58% of the documents produced nationwide are carried out with affiliated academics in the same center and 42% are carried out by academics from different centers, but always within the same country.
Through scientific mapping, it is identified that co-authorship networks among academics are endogamy, that is, they are isolated from each other (Figure 4).  Table 4 shows the ranking of institutions that concentrate a greater number of affiliations. James Cook University (Australia) leads the ranking with 12 affiliations, followed by the University of Las Palmas de Gran Canaria, with 10 affiliations. Wadden Academy-KNAW Netherlands 6 6 Note: R = ranking; A = authors; As = authorships. Source: Own elaboration.
The collaboration networks that are generated within this thematic approach are built under two criteria: by geographical origin, where it is identified that 70.9% of the documents are carried out by academics residing within the same country and only 29.1% in collaboration with academics from different countries. The second criterion is by institutional affiliation, 100% of the documents produced in multiple collaboration at international level are by academics who are affiliated in centers located in different countries, whereas 58% of the documents produced nationwide are carried out with affiliated academics in the same center and 42% are carried out by academics from different centers, but always within the same country.
Through scientific mapping, it is identified that co-authorship networks among academics are endogamy, that is, they are isolated from each other (Figure 4).

Citations
In an initial citation analysis, it can be seen that throughout the 35 years of production evolution, the 185 documents indexed in Scopus have 2091 citations with a ratio of 11.30 citations/article. The Hirsch Index is 22, which means that at least 22 articles of the total documents have received 22 or more citations. 12% of the documents reach the citation threshold. 2010 is the year with the highest number of citations, with 15% of the total (317 citations). Table 5 shows the most cited documents and it is observed that there are documents of 10 years or more. The absence of articles from recent years is due to the fact that they have not reached the required dissemination to be consolidated as references of the subject, a fact that limits the amount of citations they can receive [77].
The three most cited articles are: Cultural rural tourism: Evidence from Canada by MacDonald and Jolliffe [78], with 141 citations, followed by Implementing std on a small island: Development and use of sustainable tourism development indicators in Samoa by Twining-Ward and Butler [79], with 125 citations and finally, The sustainability of island destinations: Tourism area life cycle and teleological perspectives. The case of Tenerife by Oreja Rodríguez et al. [25], with 106 citations. Although the first two documents in the ranking accumulate 13% of the total citations, the paper by Oreja Rodríguez et al. [25] has the highest average citations/years.
On the other hand, the co-citation analysis reveals the frequency with which the authors of the previous literature are cited jointly by authors of subsequent literature. These groups generate clusters, which have a central node that is distinguished by size, that is, the volume reached by the shape that represents it determines the co-citation trend reached by each author (Figure 5). Table 5. Ranking of the most cited articles.

Authors
Title Year Donald and Jolliffe [78] Cultural rural tourism: Evidence from Canada 2003 1 ng-Ward and Butler [80] Implementing std on a small island: Development and use of sustainable tourism development indicators in Samoa 2002 1 dríguez, Parra-López and Yanes-Estévez [25] The sustainability of island destinations: Tourism area life cycle and teleological perspectives. The case of Tenerife 2008 1 rrell and Runyan [79] Ecology and tourism 1991 z, Montes, Rodriguez and Tapia [81] Rethinking the Galapagos Islands as a complex social-ecological system: Implications for conservation and management 2008 Thur [82] User fees as sustainable financing mechanisms for marine protected areas: An application to the Bonaire National Marine Park 2010 ry-Borg and Scarpa [83] Valuing quality changes in Caribbean coastal waters for heterogeneous beach visitors 2010 s, Graci and Holmes [84] Does the tourist care? A comparison of tourists in Koh Phi Phi, Thailand and Gili Trawangan, Indonesia 2010 , Watkinson and Côté [85] Managing dive tourism for the sustainable use of coral reefs: Validating diver perceptions of attractive site features 2009 eniuk and Rothley [86] Costs    (27) and Hawking (25). At this point, it is necessary to mention that 70% (19) of the authors are not represented within the database under review. Their presence within the scientific mapping is due to conceptual contributions in relation to sustainable development, ecotourism and sustainable tourism. Each cluster in Figure 5 forms a "school of thought" [42,65,66], which allows us to observe the approaches shared between the authors. Cluster 1, which is violet, is composed of 2 academics that address traditional ecological knowledge, including Folke (37 co-citations) and Berkes (23). This school of thought focuses on linking social and ecological mechanisms to build resilience within human groups [87,88] and on the use of traditional ecological knowledge for resource management [89][90][91].

Journals
A total of 118 journals were identified. The most productive ones are Ocean and Coastal Management and Journal of Sustainable Tourism, with nine articles each. The latter has the highest number of citations received by accumulating 286 (Table 6).
A total of 94 of the total number of journals only publish one article, while the remaining 24 journals publish two or more. In relation to the geographical origin of the journals, 32.20% (38) were published mainly in the United Kingdom, followed by the United States, with 14.41% (17) of the total resources.
The quartile analysis shows a high inference evaluation on the quality of the documents produced within the thematic approach, that is, 8 out of the 10 most cited journals are classified in Q1, one in Q2 and one does not have the quartile calculation in Scopus yet. This distribution suggests that the thematic approach is being published in resources of high impact and quality, which is an approximate variable to measure the research quality ( Figure 6). A total of 94 of the total number of journals only publish one article, while the remaining 24 journals publish two or more. In relation to the geographical origin of the journals, 32.20% (38) were published mainly in the United Kingdom, followed by the United States, with 14.41% (17) of the total resources.
The quartile analysis shows a high inference evaluation on the quality of the documents produced within the thematic approach, that is, 8 out of the 10 most cited journals are classified in Q1, one in Q2 and one does not have the quartile calculation in Scopus yet. This distribution suggests that the thematic approach is being published in resources of high impact and quality, which is an approximate variable to measure the research quality ( Figure 6).   The co-citation analysis of resources identifies the existence of two clusters (Figure 7). The first green cluster consists of 4 resources and collects the resources with the highest number of shared citations (between 253 and 42 citations), with the central node being Annals of Tourism Research, with 253 shared citations. The second red cluster, which consists of 10 resources, is made up of the resources with the least amount of shared citations (between 58 and 30 citations), with the central node being Ambio, with 58 shared citations. The co-citation analysis of resources identifies the existence of two clusters (Figure 7). The first green cluster consists of 4 resources and collects the resources with the highest number of shared citations (between 253 and 42 citations), with the central node being Annals of Tourism Research, with 253 shared citations. The second red cluster, which consists of 10 resources, is made up of the resources with the least amount of shared citations (between 58 and 30 citations), with the central node being Ambio, with 58 shared citations.   Figure 8 shows the need to establish whether the Law of Bradford [118] is developed within this thematic approach, that is, to examine the production and identify whether a high percentage of studies are published in a small number of journals. The Minimum Bradford Zone (MBZ) at 47, which is a value that helps determine the Bradford core by identifying the group of journals that add a descending productivity equal to 47. It is observed that the MBZ is made up of 7 journals. Figure 8 shows the inequality generated in the publication trend of the articles in the resources identified within the review (Lorenz curve).
Sustainability 2020, 12, x FOR PEER REVIEW 16 of 27 Figure 8 shows the need to establish whether the Law of Bradford [118] is developed within this thematic approach, that is, to examine the production and identify whether a high percentage of studies are published in a small number of journals. The Minimum Bradford Zone (MBZ) at 47, which is a value that helps determine the Bradford core by identifying the group of journals that add a descending productivity equal to 47. It is observed that the MBZ is made up of 7 journals. Figure 8 shows the inequality generated in the publication trend of the articles in the resources identified within the review (Lorenz curve).

Thematic Areas
The thematic areas in which the resources are classified are shown in Table 7. It is observed that the area of knowledge of Social Sciences is the first one, with 73 articles and 39 journals, followed by Environmental Science, with 31 articles and 17 journals. A varied classification of resources such as Agricultural and Biological Sciences, Arts and Humanities, Business, Management and Accounting,

Thematic Areas
The thematic areas in which the resources are classified are shown in Table 7. It is observed that the area of knowledge of Social Sciences is the first one, with 73 articles and 39 journals, followed by Environmental Science, with 31 articles and 17 journals. A varied classification of resources such as Agricultural and Biological Sciences, Arts and Humanities, Business, Management and Accounting, Medicine, Computer Science, Energy, among others; determines the multidisciplinary nature of the approach.

Keywords
Despite the current relevance in the use of keywords within several analyses, 29 documents are identified during the review, which do not have this section. In the rest of the documents, the co-occurrence analysis of keywords is applied. The terms with a high coincidence (Table 8) are tourism (20), sustainable tourism (20), ecotourism (16), sustainable development (10) and sustainability (9). Cultural tourism 6 6 10 Coral reefs 5 5 11 Small island 5 4 Note: R = ranking; f = frequency. Source: Own elaboration. Figure 9 shows the scientific mapping of the co-occurrence analysis of keywords, showing current or recent thematic nodes that are categorized as topics of interest. These are identified in yellow, highlighting topics such as "small islands", cultural tourism, biodiversity, development, sustainability, coral reefs and tourism activities such as scuba diving. Figure 9 shows the scientific mapping of the co-occurrence analysis of keywords, showing current or recent thematic nodes that are categorized as topics of interest. These are identified in yellow, highlighting topics such as "small islands", cultural tourism, biodiversity, development, sustainability, coral reefs and tourism activities such as scuba diving.

Bibliographic Analysis
The bibliographic analysis follows the analysis methodology of Álvarez-García et al. [119], where the documents are organized in (Table 9): (a) Line of research which the study is aimed at 7 subcategories: carrying capacity; responsible environmental behavior; willingness to pay; destination management: resources and new

Bibliographic Analysis
The bibliographic analysis follows the analysis methodology of Álvarez-García et al. [119], where the documents are organized in (Table 9): (a) Line of research which the study is aimed at 7 subcategories: carrying capacity; responsible environmental behavior; willingness to pay; destination management: resources and new products; impacts due to tourism development; perceptions about the local economy and willingness to visit). natural, cultural/natural and cultural resources, but it also includes an additional one that groups the documents that address the study of both natural and cultural resources). (d) Objective, the process of applied study is detailed (3 subcategories: information analysis documents, documents that develop proposals and eminently theoretical documents). (e) Sources used, the type of sources from which the data used in the study are extracted are clarified (3 subcategories: primary sources, secondary sources and use of both primary and secondary sources). (f) Brief description of the research line.  The perceptions about the economy are investigated based on the exploitation given to the different resources that host insular spaces.
Willingness to visit The willingness that tourists show to visit tourist areas in which specific and unique resources are exploited within island territories is studied.
Note: A = articles; N = natural resource; C = cultural resource; C/N = cultural/natural resource; N/A = both types of resources; IA = information analysis; D = development of proposals; T = theoretical; 1 = primary sources; 2 = secondary sources; 1y2 = primary and secondary sources. Source: Own elaboration.

Conclusions
The articulation of a systemic analysis, which consists of a bibliometric and bibliographic study, allows this research to be a consultation tool for researchers by providing a thorough mapping of the literature related to use the cultural and natural resources by tourism in island ecosystems.
The scientific production was made up of 185 documents identified in the international Scopus database. The bibliometric analysis determines that the thematic approach has been developed for 35 years, beginning with the first work in 1985. The years with the highest production are 2013 and 2017, with 18 articles each, while the year with the highest number of citations is 2010, with an accumulation of 15% of the total (317 citations).
The citation analysis shows that over 70% of documents receive between 1 and 24 citations. On the one hand, the ACA demonstrates the presence of co-citation of authors external to those identified in the review base due to the influx of studies that are inclined towards exploitation regarding sustainable development, ecotourism and sustainable tourism. On the other hand, the co-citation analysis of journals shows the 10 resources with the highest co-citation trend (see Table 6), being therefore the main publication resources on this subject.
With respect to production by authors, there is a predominance of transient researchers, with a low participation of aspiring authors, which indicates the youth of the subject. The co-authorship relationship is established as 2.65 authors/article, peer production is predominant. Production based on geographical affiliation has a clear Anglo-European dominance, led by the United States and Spain. In relation to institutional affiliation, universities predominate by concentrating 71.6% of affiliations, with the registration of affiliations to public sector units in different countries being novel, as well as to several internationally renowned museums. James Cook University based in Australia is the center with the highest number of affiliations.
Dispersion of articles is established based on Bradford's core, which raises considerable inequality by showing that 25% of review documents are published in 6% of the journals. The journals with the highest concentration of publications are Ocean and Coastal Management and Journal of Sustainable Tourism with 9 publications, but Annals of Tourism Research is the journal that reaches the highest co-citation, with 253 shared citations. Social Sciences predominates in the classification area. In addition, it is observed that of the top 10 most published journals, 80% are from quartile Q1 in the Scimago Journal and Country Rank. In the case of keywords, a correct use of keywords within studies can become a complex task to achieve, which in many cases is poorly recognized. The most co-occurring keywords are tourism, sustainable tourism, ecotourism, sustainable development and sustainability, which shows that there are new hot topics in the study trend.
The bibliographic analysis identifies 7 lines of research, the most important are destination management: resources and new products. Most studies use primary sources and there are very few studies that address the study of the use of cultural/natural resources by tourism developed in an environment such as the islands.
These results are coincident with bibliometric analyzes performed in different types of tourism; creative tourism [119], active tourism [120], adventure tourism [121], community-based tourism [122]; wine tourism [123], rural tourism [124], all these types of tourism depend on cultural and natural resources. Because they are specific types of tourism, the interest of researchers is lower than when tourism in general is investigated.
In summary, there is a growing interest in the subject in the last decade, but this does not end up consolidating. The majority of researchers are transients with a single document, not continuing with research in the area and therefore, not reaching an adequate level of specialization. This implies a small number of researchers on the subject and little research focused on tourism based on cultural and natural resources on island ecosystems. Research is emerging, there is an important niche, as well as a gap in the scientific literature that should be developed in the future.
The production trend line indicates that research will grow in the coming years. The topics discussed and included in Table 9 should continue to be addressed as they are incipient lines of research. However, there are many lines of research followed by researchers in studies in tourism in general or in other types of tourism that should be addressed by this theme (tourism based on the use of cultural and natural resources) in island ecosystems. Given the multidisciplinary nature of tourism, this involves a large number of areas of knowledge, the issue must be addressed from all of them (Business and Management, Sociology, Psychology, Geography, History, Education).
Regarding the topics to be addressed, the studies by Sáez et al. [125] and Koseoglu et al. [126] are taken into account. This latest research includes studies that suggest new frameworks on research topics and agendas such as that carried out by Ashworth and Page [127], Kandampully et al. [128], Kim et al. [129], Law et al. [130], Morrison [131], Tracey [132]. The topics to be addressed from the Business and management discipline are: Marketing: quality of service, research in marketing-mix variables such as price or communication, ethics and social responsibility [126,133], customer loyalty [128,134], consumer behavior [129,[135][136][137], image of destiny [138].
The first limitation of this research refers to the fact of focusing on a single database. Although Scopus together with Web of Science (WoS) are the two most important databases of an international scope, there are others of less relevance (Scielo, Latindex, Science Direct, Emeral, etc.). This implies that not all the scientific literature on the subject is included in the bibliometric analysis. The second limitation stems from the use of a given search equation. Although it was defined taking into account the greater number of relevant terms in the field of study, it cannot be said that they are all included.
Author Contributions: All authors contributed equally to this work. All authors wrote, reviewed and commented on the manuscript. All authors have read and agreed to the published version of the manuscript.